Arxiv Paper Reviews

Interact with arXiv Crawler API to fetch papers, read reviews, and submit comments. Use when working with arXiv papers, fetching paper lists by date/category/interest, viewing paper details with comments, or submitting paper reviews via API at http://150.158.152.82:8000.

What to check before installing/use: - Verify the remote API: The skill uses an IP address (http://150.158.152.82:8000). Confirm who runs that server and whether you trust it. Prefer a TLS-secured domain (https) and known operator. - Inspect the code locally: paper_client.py and install-deps.sh are small and readable; open them yourself before running. The client only reads config.json and makes HTTP calls. - Run in an isolated environment: use the included virtualenv (install-deps.sh) or a container to avoid affecting your system. - Protect secrets: apiKey is optional and stored in config.json; don't put sensitive credentials there unless you trust the server. Avoid posting personal/identifying info as comment content. - Posting risk: the skill can submit comments to the remote service. If you will auto-generate comments (LLM integration), ensure content is appropriate and rate limits are respected. - Metadata inconsistency: registry claimed 'instruction-only' but code files exist — treat the package as code-containing and inspect before executing. If you cannot verify who operates the API or the server's purpose, do not supply API keys or submit content to that endpoint. If you want to proceed, ask the package author for ownership/hosting information or switch apiBaseUrl to a known/trusted service.

Type: OpenClaw Skill Name: weak-accept Version: 0.1.1 The skill is designed to interact with an arXiv Crawler API, making network requests to `http://150.158.152.82:8000` as specified in `SKILL.md` and implemented in `paper_client.py`. The `install-deps.sh` script safely sets up a Python virtual environment and installs the `requests` library. There is no evidence of data exfiltration, malicious execution, persistence mechanisms, or prompt injection attempts against the agent. All network and file access (reading `config.json`) is directly aligned with the stated purpose of fetching and submitting paper reviews.